Introduction

Coronaviruses (CoVs) are a family of enveloped viruses with a positive-stranded RNA genome, and they can be genetically divided into four genera (Alphacoronavirus, Betacoronavirus, Gammacoronavirus, and Deltacoronavirus) that belong to the family Coronaviridae of the order Nidovirales [1].

CoVs are distributed widely among mammals and birds [2,3,4,5]. Individual CoVs usually infect their hosts in a species-specific manner, with alpha- and betacoronaviruses mainly infecting mammals, gammacoronaviruses normally infecting avian species, and deltacoronaviruses infecting both mammals and avian species. Attention to coronaviruses has increased in recent years because of the emergence of severe acute respiratory syndrome (SARS), Middle-East respiratory syndrome (MERS), and the newly emerging coronavirus disease 2019 (COVID-19), all of which cause acute respiratory illness, with mortality rates up to 9.5% in SARS and 35% in MERS, and with an inferred infection fatality rate varying from 0.00% to 1.63%, with corrected values varying from 0.00% to 1.54% for COVID-19 [6,7,8,9,10].

Porcine deltacoronavirus (PDCoV), a member of the genus Deltacoronavirus, is a novel swine enteropathogenic coronavirus that causes acute diarrhea, vomiting, and dehydration in neonatal piglets [11,12,13,14,15]. PDCoV was initially identified in 2012 during a molecular surveillance study in Hong Kong and emerged later in swine herds with diarrhea in Ohio (USA) in 2014 [2, 16]. Subsequently, the outbreak exhibited a global spread, and the virus has been detected in fecal samples from piglets in South Korea, mainland China, Thailand, Vietnam, and Laos [17,28, 29]. Downstream of ORF1a and ORF1b, there are several additional ORFs that code for the structural proteins: spike (S), envelope (E), membrane (M), nonstructural protein 6 (NS6), nucleocapsid (N), and nonstructural protein 7 (NS7) [17, 30]. The S protein forms peplomers on the virion surface and plays an important role in receptor attachment and viral and host cell membrane fusion [31,2]. Li et al. observed that PDCoV can efficiently infect cells with an unusually broad species range, including human and chicken cells [58]. Similar results were also reported in the comparative transcriptome profiling of human and pig intestinal epithelial cells after PDCoV infection [57]. Therefore, it appears inevitable that similar zoonotic events will occur again in the future, so more epidemiological studies need to be conducted to clarify the origin, epidemiology, and interspecific transmission mechanisms of coronaviruses.

Evolution

To date, 104 complete genome sequences of PDCoV isolates from the United States, China, Korea, Laos, Vietnam, Thailand, and Japan are available in the GenBank database. To investigate the molecular origin and evolution of PDCoV in mainland China, we performed a detailed phylogenetic analysis at the genome level using strains that have been reported in countries for the first time or for which epidemiological data show a high prevalence.

A total of 22 strains of porcine deltacoronavirus and seven avian isolates from seven countries were selected for phylogenetic analysis. Of these, 14 PDCoV isolates were from 13 different provinces of mainland China (Fig. 3).

Fig. 3
figure 3

Phylogenetic analysis of the complete genome sequences of 29 members of the genus Deltacoronavirus. The tree was constructed using the distance-based neighbor-joining method in MEGA7.0. Bootstrap analysis was carried out on 1000 replicate data sets, and values are shown adjacent to the branching points. Red represents the Chinese PDCoVs, blue represents the Vietnamese, Laotian, and Thai PDCoVs, yellow represents the US, Korean, and Japanese PDCoVs, green represents the Hong Kong (China) PDCoVs, and grey represents the avian deltacoronaviruses.

Full size image

The phylogenetic analysis showed that the PDCoV isolates clustered together, while the avian deltacoronaviruses formed a separate cluster, and the evolutionary relationship between the two is distant. All porcine deltacoronavirus strains are closely related in the genetic evolution and have a high degree of sequence similarity, indicating that they might have originated from a common ancestor. The US, Korean, and Japanese PDCoV isolates grouped in the same branch with up to 99.9% nucleotide sequence identity and therefore might represent the same strain. The isolates from Vietnam, Laos, and Thailand belong to another branch with 98.4–99.8% whole-genome nucleotide sequence identity. It is clear that the Southeast Asian PDCoV isolates are much more closely related to the Chinese strains.

The nucleotide sequence identity of the 14 strains from mainland China ranged from 97.7% to 99.7%, with the isolates from Qinghai and Gansu and the Anhui strains belonging to the same branch with 99.1–99.5% nucleotide sequence identity. The isolates CHN/AH/2004, CHN/GS/2016, and CHN/QH/2017 were found to be more closely related to HKU15-44, while the Jiangsu isolate (CHN/Jiangsu/2014) was more closely related to HKU15-155.

Compared with the US, Korean, and Japanese PDCoV strains, most Chinese strains (except HKU15-44, CHN/AH/2004, CHN/GS/2016, and CHN/QH/2017) have a continuous deletion mutation of three nucleotides (AAT) at one site in the S gene. Previous reports have shown that the mutation rate of the S gene is relatively high, which may lead to altered tissue tropism, virulence, and even host specificity [27]. Whether the deletion of ATT has an effect on the virulence of the virus needs to be studied further.

In general, analysis of genetic evolution based on whole-genome sequencing shows that PDCoV has undergone extensive variation in different regions, and the mutations occur mainly in the S gene. Thus, it is important to monitor genetic variations occurring in the PDCoV S gene as well as to evaluate the impact of these variations on pathogenicity in order to develop an effective vaccine to control the disease.

Virulence and pathogenicity

Coronavirus infection has been documented previously in livestock and companion animals [55, 91]. TGEV and PEDV and the newly reported SADS-CoV mainly cause severe intestinal infections in piglets, leading to high morbidity and mortality and vast economic losses [88, 92, 93]. Bovine CoV, rat CoV, and infectious bronchitis virus (IBV) cause mild to severe respiratory tract infections in cattle, rats, and chickens, respectively [94,95,96]. Feline infectious peritonitis virus (FIPV) causes highly lethal disease in domestic cats [97].

Clinically, PDCoV can cause infection in pigs of various ages but mainly causes infection in newborn piglets, characterized by mild to severe diarrhea, vomiting, dehydration, anorexia, and growth retardation [13, 15, 98]. Inoculation experiments have suggested that although PDCoV exhibits enteropathogenicity in both gnotobiotic and conventional piglets, infected pigs display milder signs of clinical impact and disease severity than those infected with PEDV and TGEV [12, 24, 82].

Due to their underdeveloped immune system, neonatal piglets are highly susceptible to viral infection during their first few weeks. Mortality rates are highest in neonatal piglets, often reaching nearly 100%. Commercial fattening pigs and sows can also exhibit typical clinical features, such as diarrhea, inappetence, and persistent viral shedding in feces, but the symptoms of fattening pigs and sows are relatively mild, and the mortality rate is lower, with the animals gradually recovering [15].

Within 20–48 h postinfection, diarrhea is observed. Diarrhea typically lasts for at least 1 week. Vomiting symptoms were inconsistent in different experiments, which may be due to differences in virulence between strains. Core body temperatures remained within normal limits, and no respiratory signs were observed. Viral shedding peaked on day 7 postinfection, and virus was still detectable in feces and in the ileum at day 21 postinfection, which may enhance the risk for viral transmission [12,13,14, 21, 24, 52].

Pathological changes are characterized by intestinal villous atrophy and shortening, and villous changes are associated with extensive intestinal epithelial degeneration and necrosis. Gross lesions are observed in the small intestines, and no significant lesions are observed in extraintestinal tissues except the lung. PDCoV infection can cause mild interstitial pneumonia in gnotobiotic piglets [12], which has not been reported for PEDV or TGEV.

Vaccine and control strategies

Vaccines remain the most effective means to control coronavirus infections. However, there are no effective vaccines available for PDCoV. Strategies for PDCoV vaccine development include inactivated virus vaccines, subunit vaccines, viral vector vaccines, and live-attenuated virus vaccines, each of which has both advantages and disadvantages. Multiple routes of vaccine research are being evaluated for PDCoV and other enteric coronaviruses.

Inactivated virus vaccines

Inactivated virus vaccines use chemicals or radiation to render the virus noninfectious while preserving its antigenicity. The most recent PDCoV vaccine was developed by the State Key Laboratory of Veterinary Biotechnology in China. The vaccine is based on inactivated virus formulated with an adjuvant. When administered to seronegative sows using a prime/boost strategy 20 and 40 days before delivery, high levels of spike (S)-specific IgG and neutralizing antibody against PDCoV were found in colostrum and milk, as well as in the serum of piglets born to vaccinated sows [12]. Piglets were infected orally at 5 days of life with 105 TCID50 of PDCoV. The experiment showed that 87.1% of all piglets (n = 31) born to immunized sows were protected against lethal infection, and the infected piglets showed milder diarrhea, less viral shedding, and only minor damage to intestinal villi. In contrast, piglets from unimmunized sows had moderate diarrhea, which quickly worsened at 2 days postinfection and remained severe until the end of the experiment.

Live-attenuated vaccines

Nonreplicating vaccines (inactivated vaccines, subunit vaccines) usually generate short-lived neutralizing antibody responses with comparatively low titers. In contrast, live-attenuated vaccines are generally more immunogenic than nonreplicating vaccines; they can induce long-lasting immunity, produce a comprehensive spectrum of native viral antigens, and present antigens to the immune system in the same manner as in natural infections.

Zhang et al. [99] generated a full-length infectious cDNA clone of PDCoV, which they manipulated by replacing the NS6 gene with a green fluorescent protein (GFP) to generate rPDCoV-ΔNS6-GFP. Growth kinetics studies suggested that rPDCoV-ΔNS6-GFP showed a substantial reduction in viral replication in cell cultures and was highly attenuated in neonatal piglets, indicating that PDCoV lacking NS6 might be an ideal live-attenuated vaccine candidate.

Generally, live-attenuated virus vaccines are promising candidates for use against coronavirus infections, but they also have decreased safety and stability compared to inactivated vaccines, and some live-attenuated virus vaccines have the potential to spontaneously revert to virulence post-vaccination.

Vectored vaccines

Vectored vaccines function as viral gene delivery systems that rely on a host viral genome from a different virus, such as adenovirus, poxvirus, measles virus, parainfluenza virus, rabies virus, or vesicular stomatitis virus, and they have been used in the development of vaccines for CoVs [100,101,102,103,104,105,106].

Porcine adenovirus was used to deliver the core neutralizing epitope of PEDV, and this resulted in robust humoral and mucosal immune responses in piglets [100]. A recombinant vesicular stomatitis virus expressing the PEDV spike protein was developed, and sows immunized with this recombinant vaccine provided protective lactogenic immunity against a virulent G2b PEDV challenge to their piglets [104]. In addition, Yuan et al. used swinepox virus to express an epitope of the S protein of TGEV and a truncated spike protein of PEDV [101, 102].

Virus-like particles (VLPs)

Virus-like particles (VLPs) have drawn increasing attention in recent years. VLPs containing one or more viral structural proteins structurally resemble the native virus, can be easily recognized by antigen-presenting cells and B cells, and are capable of eliciting robust humoral and cell-mediated immune responses that are comparable to those achieved with inactivated or live-attenuated virus vaccines. There have been few reports about PDCoV VLPs, but studies of other animal coronavirus VLPs can provide a reference for PDCoV vaccine research.

Wang et al. [107] produced PEDV virus-like particles (VLPs) composed of S, M, and E proteins using a baculovirus expression system and showed that they induced a high level of anti-PEDV-neutralizing antibodies in mice. Xu et al. [108] developed chimeric IBV VLPs expressing M, E, and a recombinant S protein in baculoviruses. These induced a high level of IBV-specific antibodies and neutralizing antibodies that were comparable to those induced by an inactivated M41 virus via subcutaneous inoculation.

Moreover, other vaccine approaches for expressing the S, E, M, and N genes of two or more coronaviruses as well as other viral genes in bacteria, yeast, plants, and nanoparticles have been assessed [109,110,111,112,113,114,115,116]. However, the efficacy of these VLPs against lethal infection has not been tested in piglets, and further studies need to be performed.

Currently, most commercial vaccines for enteric coronaviruses are designed to induce lactogenic immunity by vaccinating the sow during pregnancy, and antibodies are passively transferred from sows to neonatal piglets via colostrum and milk.

Coronavirus infections are generally initiated at mucosal surfaces, and it is critical to induce localized intestinal sIgA and T cell immune responses to mucosal infections. For maternal immunity, oral vaccines or intentional infection of the sow may initiate the gut-mammary sIgA axis [117, 118].

A previous study showed that oral inoculation of sows with attenuated TGEV, followed by intramuscular injection with a recombinant subunit vaccine expressing the S protein of TGEV as a booster generated high titers of sIgA antibodies and neutralizing antibodies in colostrum and milk [119]. Similar prime/boost strategies can be applied to PDCoV vaccines to induce active immunity in newborn piglets.

Coronaviruses are an important group of pathogens that can have a devastating impact on humans and animals. New zoonotic coronaviruses are continually emerging or reemerging. In addition to good production management and strict biosecurity measures, the most effective way to control PDCoV is vaccination. Consequently, new vaccine development platforms and technologies are highly desirable, and further research will provide a better understanding of PDCoV replication and pathogenesis, a prerequisite for the development of new and promising vaccines to prevent, control, and ultimately eliminate the virus.